Understanding How Value Neurons Shape the Generation of Specified Values in LLMs
This work addresses value alignment concerns in LLMs by providing a method to interpret and control internal value representations, which is incremental as it builds on existing interpretability techniques.
The authors tackled the problem of opaque value representations in large language models by introducing ValueLocate, a mechanistic interpretability framework that localizes value-critical neurons using a dataset based on the Schwartz Values Survey, and demonstrated that manipulating these neurons alters model value orientations.
Rapid integration of large language models (LLMs) into societal applications has intensified concerns about their alignment with universal ethical principles, as their internal value representations remain opaque despite behavioral alignment advancements. Current approaches struggle to systematically interpret how values are encoded in neural architectures, limited by datasets that prioritize superficial judgments over mechanistic analysis. We introduce ValueLocate, a mechanistic interpretability framework grounded in the Schwartz Values Survey, to address this gap. Our method first constructs ValueInsight, a dataset that operationalizes four dimensions of universal value through behavioral contexts in the real world. Leveraging this dataset, we develop a neuron identification method that calculates activation differences between opposing value aspects, enabling precise localization of value-critical neurons without relying on computationally intensive attribution methods. Our proposed validation method demonstrates that targeted manipulation of these neurons effectively alters model value orientations, establishing causal relationships between neurons and value representations. This work advances the foundation for value alignment by bridging psychological value frameworks with neuron analysis in LLMs.